A New Software Predicts The Onset Of Psychosis By Analyzing The Language

This innovation can be of great help to implement mental health interventions as soon as possible.


Is it possible to diagnose a serious mental disorder several years after the first symptoms appear ? If so, what implications would it have both for the clinic and for the person at risk of developing them?

In a recently developed study, speech analysis software was used to try to predict the risk of psychosis. Throughout this article we will talk about this new computer program that is still under investigation and development.

What is psychosis?

In the field of psychology and psychiatry, psychosis refers to a state of mental alteration or a serious mental disorder through which the person experiences a loss of contact with reality and an impairment of normative social functioning.

This disorder refers to a series of serious disorders, in which the person may manifest hallucinations of all kinds, as well as delusions, alterations in personality and disorganized thinking and language.

In addition, it is possible that this symptomatology is accompanied by unusual or unusual behaviors in the person, as well as by problems when establishing and maintaining any type of interaction with other people, obstructing and hindering the day-to-day life of the patient. person.

It is necessary to emphasize that psychosis is not a single mental disorder; Rather, it encompasses any classification of psychiatric conditions known as psychotic disorders. These mental disturbances include:

  • Schizophrenia.
  • Delusional disorder.
  • Schizophreniform disorder.
  • Schizoaffective disorder.
  • Shared psychotic disorder.
  • Substance-Induced Psychotic Disorder.
  • Psychotic disorder due to a medical illness.

What is this software about?

As mentioned above, disorganized thinking is one of the main symptoms of psychosis. This, which is externalized through disorganized speech of the patient, can be evaluated through oral clinical interviews with the patient.

This disorganized thinking distinguishes because the patient’s speech becomes tangential and a loss of associations of the ideas expressed. Therefore, the patient can express a confused and incoherent speech, in addition to continuously jumping from one idea to another.

Following these events, a study conducted in collaboration with the New York Psychiatric Institute, Columbia University Medical Center and the IBM TJ Watson Research Center, evaluated and compared the transcripts of a series of interviews conducted with at-risk youth to develop a psychosis.

During the study, the interviews of 34 New York participants and 59 participants from the city of Los Angeles were evaluated, using speech analysis software, for whom the onset of psychosis was estimated throughout the two subsequent years.

After analyzing the interviews, it was found that the software was able to correctly differentiate between young people who had a real risk of developing psychosis and those who did not, in most cases.

According to experts in the analysis of the pathological behavior of people, 1% of the population between the ages of 14 and 27 is at serious risk of developing psychosis. These people tend to manifest symptoms such as unusual thoughts, tangential speech, perceptual disturbances, and suspicious reactions to other people.

Approximately 20% of this population at risk will experience a psychotic episode; Therefore, early detection accompanied by psychological intervention could help reduce the impact of this episode or even prevent it.

Speech as the key to detection

Within psychology and psychiatry, both language and behavior provide a large amount of key information in the diagnosis of any type of mental disorder.

In the case of psychoses, the oral language of the patient is a great source of information about what is happening in the person’s mind. Providing important clues about what the person experiences, feels and perceives.

In the study referenced above, patients were asked to narrate their subjective experiences. These conversations were then transcribed and analyzed using speech evaluation software.

According to the researchers, this software analyzed the expression patterns, semantics and syntax used by the participant. As well as the length of the sentences and the correct use of the words used to link the expressions.

It is necessary to specify that a clinical psychiatrist or mental health professional is able to recognize these signs or symptoms manually, through a face-to-face interview with the patient. However, the development of this software implies an increase in precision by taking into account more variables and in a more rigorous way.

What were the results of the study?

After conducting the interviews and analyzing them, the participants were followed up for two and a half years.

The examiners who developed this study found a series of qualities or key traits in the conversations that predicted future mental problems in 83% of the cases.

Specifically, those features of oral language that helped predict the onset of psychoses were alterations in the flow of meaning between sentences. Especially highlighting sentences shorter than usual and with much less elaboration.

After these results, the researchers place their hope that, thanks to this software , early and reliable detections of those more subtle symptoms that may appear even many years before the development of the psychotic disorder can be made.

Implications of this software in clinical practice

Within the field of clinical diagnosis within psychology and psychiatry, the possibility of using a computer program that helps and facilitates the early detection of serious mental disorders is very good news.

In this way, the work of diagnosis and prognosis of serious mental conditions can be streamlined, and thus design proposals for treatment and early intervention that reduce the impact of symptoms. In addition, other advantages that the researchers assign to this software is that said analysis is carried out quickly, safely and with very low economic costs.

However, the researchers themselves warn that it is only a prototype and that research time is still required to turn it into a completely reliable diagnostic tool.

Add a Comment

Your email address will not be published. Required fields are marked *